6 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. Ill 



figure 2. Thus, the wood specimens were cross-dated with one 

 another ; that is, growth layers taken to be equivalent in time were set 

 in line with one another. 



4. The thicknesses of the growth layers were measured to hun- 

 dredths of a millimeter by means of a measuring microscope. These 

 measurements are called raw data in millimeters. 



5. The average thicknesses of the growth layers on the sequences 

 HPC 4, 8, and 9 were corrected downward to approximate the 

 averages of the other sequences. Otherwise, if a sequence of high 

 average thickness were one of several merged into a group, its high 

 average would unduly influence the values in the group. 



6. The raw data in millimeters of each sequence were changed 

 into percentages of the sequence mean in order to establish an identity 

 of units and an identity of base line between tree growth and rainfall. 



7. The raw percentages were smoothed by the formula 



a + 2b + c 

 4 



8. Various sequences were merged into groups and smoothed. 

 Groups. — The nine sequences divided themselves geographically 



into three groups — east of the Pass, on the Pass, and west of the 

 Pass. Nevertheless, other groupings were arranged in order to make 

 the tests not only as critical but also as thorough as possible so far 

 as comparison with rainfall was concerned. 

 The following groups were set up : 



Group I (G i). Trees 1-5, 8, and 9. Variable and fairly variable sequences 

 only. 



Group 2 (G 2). Trees 1-3, 5, and 9. Variable sequences only. 



Group 3 (G 3). Trees 1-3, and 9. Most variable sequences based upon a 

 visual study of the wood samples. Douglas fir omitted. 



Group 4 (G 4). Trees 1-4. East of the Pass. 



Group 5 (G 5). Trees 5 and 6. On the Pass. 



Group 6 (G 6). Trees 8 and 9. West of the Pass, exclusive of the tree 

 whose sequence is uniform. 



Group 7 (G 7). Trees 1-9. All trees. 



Group 8 (G 8). Trees i, 2, and 9. Ponderosa pines with variable sequences. 



Group 9 (G 9). Trees 7-9. West of Pass. 



Group 10 (G 10). Trees 1-3, and 7. From the drier sites. 



Group II (G 11). Trees 4-6, and 9. From the wetter sites. 



The primary groups are numbers 4, 5, 9, 10, 11, and 7. 



Selection of rainfall stations and the treatment of data. — A moun- 

 tainous country permits little choice in the selection of rainfall stations. 

 Fortunately, one station, Chacon, lies approximately 7 miles, airline 



